The proposed research seeks (a) to understand the dynamics of inner-city black junior high school students' behaviors that place them at risk for sexually transmitted HIV infection and (b) to discover ways of reducing the frequency of such behaviors. The research will assess black adolescents' knowledge of AIDS and other sexually transmittable diseases (STD) and their attitudes, beliefs, and self-efficacy regarding safe sex practices and relate them to risky sexual behaviors. It will also implement and evaluate an intervention designed to reduce the adolescents' tendency to engage in risky sexual activities. Participants will be 576 black junior high school students (7th and 8th graders), who will be randomly assigned to the conditions of a 3 x 3 x 2 (Intervention Type x Sexual Composition x Sex of Facilitator) factorial design. Intervention type contrasts (a) a one-shot AIDS prevention group that receives an intervention designed to increase knowledge of STD, including HIV infection, to modify maladaptive attitudes, to increase skill at negotiating difficult sexual situations, and to heighten motivation to implement safe sex practices; (b) a two-shot AIDS prevention group that receives the AIDS intervention in Year 1 of the study and again in Year 2; and (c) a control group that receives a two-shot general health promotion intervention. Sexual composition contrasts students who receive interventions in small all-male, all-female, or mixed- sex groups. Sex of facilitator contrasts students who receive interventions in groups led by a male facilitator or by a female facilitator. We will evaluate the effects of the interventions on the frequency of risky sexual behaviors, on correlates of such behaviors, including self-reported STD and pregnancy, and on factors hypothesized to influence risky behaviors. These factors include not only sexual knowledge, but also attitudes, perceived norms, and intentions regarding risky sexual activities and self- efficacy regarding safe sex practices--mediators of behavior change suggested by the theory of reasoned action and social learning theory. Baseline data will be collected prior to interventions, and we will evaluate the interventions' effects on measures collected at 1 month, 4 months, 8 months, 1 year, 2 years, and 3 years post intervention. Baseline analyses will seek to determine whether correlates of risky sexual behavior suggested by prior research and theory are evident in this population; prospective analyses will seek to identify early adolescent (7th and 8th grade) precursors of later adolescent (10th and 11th grade) sexual activity, knowledge, attitudes, and self- efficacy. We will examine whether the quality of the small group discussions mediates the effects of interventions. We will also examine factors that might interact with interventions--whether effects of interventions differ among adolescents who differ on key dispositional variables. Such potential moderators include sexual anxiety, assertiveness, impulsivity, and cognitive development. Several statistical procedures, including mixed- model analyses of variance and hierarchical multiple regression, will be used.